# High-precision classification

PP-LCNet_x1_0_table_cls is an efficient table classification model used to classify input table images and supports the classification of lined tables and borderless tables.

Document image orientation classification module, used to distinguish the orientation of document images and correct them through post-processing to improve the accuracy of OCR processing.

Multilingual E5 large pooled model supporting sentence similarity calculation and feature extraction tasks in multiple languages.

A vision-language encoding model fine-tuned based on the SigLIP 2 architecture, specifically designed for multi-category image classification, particularly suitable for filtering sensitive, suggestive, or security-related media.

A vision Transformer model fine-tuned specifically for NSFW content rating, capable of identifying neutral/mild/moderate/severe four-level content risks

This is a Vision Transformer (ViT) model fine-tuned on a legume dataset for identifying disease conditions in legume leaves.

A FocalNet fine-tuned NSFW image classification model for content moderation tasks, categorizing images into safe, needs review, and unsafe classes.

An audio classification model based on the wav2vec2 architecture, suitable for climate-related sound classification tasks

A multi-task fine-tuned model based on ModernBERT-large, specializing in Natural Language Inference (NLI) tasks, excelling in zero-shot classification and reasoning tasks.

A multi-label content moderation system built on the DistilBERT architecture for detecting and classifying potentially harmful content in user comments, featuring high accuracy and lightweight characteristics.

This model is a demonstration classifier designed to identify problematic content (wrong language, garbled text) in Danish and Swedish web pages.

AIM-v2 is an image encoder model based on the timm library, with a scale of 1 billion parameters, suitable for image feature extraction tasks.

This model is used to detect whether an image is real or AI-generated, employing the Vision Transformer (ViT) architecture for precise classification.

A text classification model fine-tuned using LoRA technology based on the e5-small model, specifically designed for high-accuracy differentiation between AI-generated and human-written text

A text classification model fine-tuned based on GPT-2, used to distinguish AI-generated text, Zhihu user responses, and texts from other sources.

Zero-shot image classification model based on OpenCLIP library, suitable for patent image analysis

A crop disease recognition model fine-tuned based on Google Vision Transformer (ViT) architecture with 70% accuracy

A large language model with 7B parameters based on Qwen2 architecture, focusing on sentence similarity calculation and text embedding tasks.

MobileCLIP-S2 is an efficient image-text model that achieves rapid inference through multimodal reinforcement training, delivering outstanding zero-shot performance while maintaining a compact size.

A stuttering detection model fine-tuned based on the DistilHuBERT model, achieving 90.24% accuracy on the evaluation set

A DistilBERT-based binary text classification model for identifying safe content and restricted (NSFW) content, suitable for content moderation scenarios.

Cutting-edge language model specifically designed for Hebrew, a BERT-based sentiment analysis model

An IndoBERT-based model for identifying sexual harassment content, used to detect related content in text.

A NSFW (Not Safe For Work) content detection model fine-tuned based on Google Vision Transformer, capable of identifying 5 types of image content

A text classification model fine-tuned based on distilroberta-base, used to identify rejection responses generated by large language models

This model was created by fine-tuning the umm-maybe AI art detector on the Wikimedia-SDXL image pair dataset, specifically designed to detect images generated by SDXL.

An X-ray image detection model fine-tuned based on AI-image-detector, achieving 99.83% accuracy

This is a sentiment classification model for internal datasets of Brazil's SEBRAE-RS institution, capable of classifying Portuguese texts into positive, neutral, or negative sentiments.

Ember v1 is an embedding model based on sentence-transformers, primarily used for feature extraction and sentence similarity calculation.

bge_micro is a lightweight sentence similarity calculation model based on transformer architecture, specifically designed for efficient feature extraction and sentence similarity tasks.

A general-purpose image classification model capable of recognizing and classifying various image contents

A Turkish hate speech detection model fine-tuned based on BERTurk, capable of identifying neutral, offensive, and hate speech texts.

Fine-tuned based on Google's ViT model for the diabetic retinopathy classification task, with an accuracy of 74.31%.

A speech processing model fine-tuned based on microsoft/wavlm-large, achieving an accuracy of 73.33% on the evaluation set

A multi-category image classification model trained on AutoTrain, focusing on solar panel object detection tasks

An image classification model based on the transformers library for identifying and classifying plant diseases.

A general-purpose image classification model capable of recognizing and classifying objects in input images

This is a binary classification model trained using AutoTrain, designed to distinguish between memes and non-meme images.

This model is an image classification model fine-tuned on the beans dataset based on Google's ViT architecture, achieving 97.74% accuracy on the validation set.

A facial expression recognition model fine-tuned based on Google's ViT model, trained on the FER2013 dataset with an accuracy of 70.22%.